Vaccines against many infectious diseases are effective because prior infection with the pathogen also induces sterilizing immunity. In contrast, there are numerous reports in the literature of disease caused by reinfection with Mycobacterium tuberculosis in those who have previously had tuberculosis, and in fact, prior tuberculosis increases the risk of subsequent disease, either through relapse or reinfection. In addition, mixed infections in tuberculosis patients have also been reported. However, seminal epidemiologic studies suggested that people with clinically latent M. tuberculosis infection can protect against development of tuberculosis due to reinfection. A review of the literature reveals that the effects of prior M. tuberculosis infection on subsequent infection are poorly studied and not well understood. In this application, we undertake a detailed study of the consequences of reinfection with M. tuberculosis. We bring together cutting edge technology for analysis of bacterial populations and immune responses in individual granulomas in an established non-human primate model of tuberculosis to address how a primary infection influences a secondary infection. We hypothesize that primary infection and the ensuing immune response will impair progression of the secondary infection at the granuloma level. We will dissect the immune responses in primary and secondary lesions to identify those factors that mediate killing of bacteria in the lesions. Finally, we will investigate clinically relevant scenarios of reinfection: challenge with a heterologous lineage of M. tuberculosis and treatment of M. tuberculosis infection prior to reinfection. The studies here will uncover basic and novel aspects of tuberculosis biology and immunology, as well as provide key translational insights into vaccine development and the consequences of treating latent infections in humans.